LA Leaderboard - PAWS-X Spanish: leaderboard
Adversarial paraphrase identification (PAWS-X) in Spanish. From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; accuracy (%), higher is better. Each model result file is a separate run.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | aya-expanse-8B | 66.95 | #914 |
| 2 | Qwen3-8B-Base | 66.9 | #626 |
| 3 | OLMo-2-1124-7B-Instruct | 66.7 | #1343 |
| 4 | Llama 3.1 8B Instruct | 66.6 | #1017 |
| 5 | occiglot-7B-eu5 | 66.6 | |
| 6 | Teuken-7B-Instruct-v0.4 | 66.6 | #1408 |
| 7 | EuroLLM-9B | 65.5 | #1238 |
| 8 | Mistral-7B-v0.3 | 65.3 | #1321 |
| 9 | Llama 3.1 8B | 65.05 | #1130 |
| 10 | Qwen 2.5 7B | 64.8 | #895 |
| 11 | Qwen 2.5 1.5B Instruct | 64.05 | #1393 |
| 12 | Teuken-7B-instruct-commercial-v0.4 | 64.05 | #1421 |
| 13 | Gemma 2 9B (IT) | 63.8 | #795 |
| 14 | Qwen 2.5 3B | 63.75 | #1177 |
| 15 | Mistral 7B Instruct (v0.3) | 63.3 | #1255 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=la-leaderboard-paws-x-spanish · How It Works · Data refreshed daily, snapshot 2026-11-05.